This article was downloaded by: [Universidad Autonoma Metropolitana]

On: 18 September 2014, At: 13:17

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Political Marketing
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/wplm20
The American Internet Voter
Thad E. Hall
a
& Betsy Sinclair
b
a
University of Utah , Salt Lake City, Utah, USA
b
University of Chicago , Chicago, Illinois, USA
Published online: 23 Feb 2011.
To cite this article: Thad E. Hall & Betsy Sinclair (2011) The American Internet Voter, Journal of
Political Marketing, 10:1-2, 58-79, DOI: 10.1080/15377857.2011.540194
To link to this article: http://dx.doi.org/10.1080/15377857.2011.540194
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
Content) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
The American Internet Voter
THAD E. HALL
University of Utah, Salt Lake City, Utah, USA
BETSY SINCLAIR
University of Chicago, Chicago, Illinois, USA
Since the creation of the Internet, there has been a seemingly
never-ending number of books and analyses about the role of the
Internet in politics. Many of these books fail to keep in mind that
the behavior of elitesthe well-educated and politically active
individuals who often represent the peer group of these authors
is not generally representative of the behavior of the public at large.
Pundits and scholars alike have remarked that 2008 appeared to
be a revolutionary year in the use of the Internet in political
campaigns, but few have systematically examined the role of the
Internet in participatory politics for the average voter. Instead of
relying on case studies, this article uses nationally representative
survey data from 2004 through 2008 to determine how the general
public usesor does not usethe Internet in their political lives.
The authors then consider whether the patterns of use for this tech-
nology appear different in the November 2008 general election
cycle. They consider this issue in the context of increasing polariza-
tion among some fraction of the American electorate and in the
policy platforms of elected officials.
KEYWORDS American voter, 2008 general election, Internet use
The authors thank the members of their panel at the 2009 American Political Science
Association Meeting in Toronto, Canada, for helpful discussion and Danielle Glazer for
research assistance.
Address correspondence to Thad E. Hall, University of Utah, 260 S. Central Campus Drive,
Room 252, Salt Lake City, UT 84112, USA. E-mail: thadhall@gmail.com
Journal of Political Marketing, 10:5879, 2011
Copyright # Taylor & Francis Group, LLC
ISSN: 1537-7857 print=1537-7865 online
DOI: 10.1080/15377857.2011.540194
58
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

INTRODUCTION
Since use of the Internet became widespread in the 1990s, there have been a
seemingly never-ending number of books, articles, and analyses about the
role of the Internet in politics. These works have identified various aspects
of the role of the Internet in the political sphere and the way in which
individuals use this tool in the political realm of their lives. For example, Marc
Ambinder wrote in June 2008 the following:
The communications revolution under way today involves the Internet,
of course, and if Barack Obama eventually wins the presidency, it will
be in no small part because he has understood the medium more fully
than his opponents do. His speeches play well on YouTube, which
allows for more than the five-second sound bites that have characterized
the television era. And he recognizes the importance of transparency and
consistency at a time when access to everything a politician has ever said
is at the fingertips of every voter . . . . Obama has truly set himself apart by
his campaigns use of the Internet to organize support. No other candi-
date in this or any other election has ever built a support network like
Obamas. The campaigns 8,000 Web-based affinity groups, 750,000
active volunteers, and 1,276,000 donors have provided him with an enor-
mous financial and organizational advantage in the Democratic primary.
The idea, of course, is that the Internet is revolutionizing politics.
1
However, the Internet could be a mechanism for accomplishing this, as sug-
gested in the quote above, in many different ways. The Web may allow indi-
viduals to be involved in politics in new ways and donate to campaigns more
easily and allow candidates to organize individuals more effectively. Interest-
ingly, the Internet revolution has occurred at the same time that there has
also been an explosion in the use of face-to-face contact in campaigns.
Numerous studies have found that such communications are very effective
in getting individuals to turn out to vote and to listen to new information
from candidates.
2
The question we are interested in examining in this article is how the
public usesor does not usethe Internet in their political lives. Both the
General Social Survey and the Pew Internet and American Life Project allow
us to gain leverage over several key questions regarding the Internet in the
political sphere and to contrast political Internet use with Internet use more
broadly. We are specifically interested in the use of the Internet by partisans
and whether there are partisan differences in Internet use patterns. We also
investigate claims that the use of the Internet by partisans could make them
more radical in their views and could make them more engaged in the pro-
cess. We examine these issues using survey data from 2004 to 2008. We find
that Internet users are not divided by a partisan difference; Democrats are not
more likely than Republicans (or other party registrants) to be active Web
The American Internet Voter 59
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

users or to use the Internet for political purpose. We see no systematic
evidence that the Internet users in 2008 are substantially different than the
Internet users in 2004, although we highlight a few small differences in our
empirical analysis. We do see indications that individuals who use the
Internet to confirm their existing political preferences are increasingly likely
to participate and additionally that users with access to the Internet are
increasingly likely to donate to political campaigns.
INTERNET AND POLITICS: THEORY AND COMMENTARY
Technological inventions have advanced the playing field for politicians for
years. The printing press, radio, television, and the Internet have provided
politicians useful outlets to get their messages out to citizens. With each
advancement, the availability of information has been expanded, making
citizens better informed and making the voting process more democratic.
Participating in the newest technologies is imperative for politicians to stay
relevant and remain competitive.
The growth of the printing press and newspaper chains made newspa-
pers a prime resource for media coverage. With literacy rates increasing and
the expansion of the printing press, politicians realized newspapers were an
ideal vehicle for speedy media coverage, making the transfer of information
more efficient and speedy. With higher literacy rates, politicians had the
opportunity to gain the attention of a larger audience. The variety of
newspaper chains helped to create competition among the newspapers for
coverage of cutting-edge stories and different editorial positions. In addition,
information could be given firsthand rather than traveling by word of mouth
from town to town (Bimber 2003; Bimber and Davis 2003).
Television entered the scene, threatening to change the entire landscape
of campaigning. It was believed to be capable of providing information to an
even broader audience than the previous technologies and bringing politics
to the masses with greater accuracy and even greater speed. It would also
give citizens an idea of the man or woman behind the politician by enhanc-
ing the transparency of political campaigns. The latest wave of technology,
the Internet, has given political campaigning a facelift and has dramatically
changed how politicians campaign. It has changed campaigning more than
the printing press, radio, and television combined. American politicians have
campaign Web sites, fundraise online, blog, e-mail citizens, and participate in
a whole host of other activities all made possible by the Internet (Cornfield
2004; Frantzich 2002; Graff 2007).
The Internet has been depicted as having the greatest capacity to improve
democracy and democratic institutions. Although the Internet is thought to
increase democratic accountability, character, and the integrity of political
campaigns, there can also be a divide between haves and have-nots (Smith,
60 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

Schlozman, Verba, and Brady 2009; Krueger 2002; Best and Krueger 2005;
Mossberger, Tolbert, and Stansbury 2003; Norris 2001). A candidates campaign
promises become a contract with the citizens (Anderson and Cornfield 2003;
Bimber 2003). The candidate, if he or she wins, can be held responsible in the
next election should he or she neglect these promises, thus promoting account-
ability. Information is more readily accessible (e.g., Kenski and Stroud 2006). A
citizen can visit one candidates Web site, review that candidates platforms, and
then visit an opposing candidates Web site and review that candidates plat-
forms. Most candidates Websites alsoallowvisitors toe-mail questions, allowing
citizens to communicate with the candidate. Another unique characteristic of the
Internet, which is something that radio and television do not offer, is that citizens
can be interactive with one another and deliberate online. All of this can be done
inexpensively, conveniently, and easily on a global scope. It increases the likeli-
hood that citizens will become more engaged and involved but can also increase
polarization (Mossberger, Tolbert, and McNeal 2008; Klotz 2004; Tolbert and
McNeal 2003; Kimball and Gross 2006; Altman and Klass 2005).
As citizens become more engaged, involved, and most importantly
tech-savvy, they develop expectations for a candidates Web site content
(Wagner and Gainous 2009; Trent and Friedenberg 2007; Chadwick 2006; Foot
and Schneider 2006). The Web site design must be visually pleasing and easy to
navigate. If a voter has a question about a specific issue, the voter wants to be
able to immediately find the answer on the candidates Web site. The Web site
cannot simply list the candidates positions. A sparse or cluttered Web site will
not gain interest and is not tapping into the full potential of the Internet. Not only
is the Web design important, but the candidates appearance on the Web
becomes important too. A candidate will want to honestly present himself or
herself and communicate the traits and accomplishments that make him or
her distinct from the other candidates. If candidates are dishonest in listing their
accomplishments, it will surface and cause skepticism. Emphasizing a common
personal background may also help to give a candidate an edge.
Once a citizen has been exposed to a candidates Web site, the candi-
dates next motives are to encourage subscription to his or her e-mail list,
move interested Web visitors from spectators to activists, receive donations,
and persuade undecided voters. E-mail lists show not only that a Web site
visitor was willing to take the time to visit the Web site but also that the visitor
was also willing to give contact information to receive highlights and updates
on the campaign, thereby broadening the supporter base (Hara and Estrada
2005; Bimber and Davis 2003; Klotz 2007). This can also lead to political dis-
cussion groups, blogging, and other online venues promoting discourse.
Too-frequent e-mails can be a turnoff to some voters; therefore, politicians
must be careful to avoid this.
The second major objective is to move Web visitors and e-mail subscribers
from passivity to activism. By mobilizing voters and getting them excited about
the campaign, a candidate can create a large volunteer group to work on the
The American Internet Voter 61
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

campaign advancing its message. E-mail reminders of rallies, candidate
appearances, and invitations help can lead to volunteerism (Bimber and Davis
2003).
In addition to e-mail lists and volunteerism, online fundraising is a
primary motivation of candidates. Raising money on the Internet poses few
obstacles. Internet donors can make a donation that is just as easy as making
an ATM transaction, and it can be done at any time of day or night. Donations
do not have to be solicited through the mail; therefore, if a voter identifies
with a candidate that voter can immediately contribute. Responses to the con-
tribution can be personally tailored and delivered instantly (Cornfield 2004).
Persuading undecided voters is also a main objective. Candidates try to
make their issue positions and biographies prominent on their home page to
lure undecided voters into clicking for more information. The difficulty
candidates face is in gauging how many of the visitors are undecided. It
becomes complicated for a candidate to know how to design the Web site
for undecided voters and decided voters too. Whether voters are undecided
or decided, they are likely to visit the Web for candidate information.
HYPOTHESES AND DATA
There are many claimssuch as the ones made aboveabout how
individuals use the Internet in politics. But how do people actually use the
Internet in major political campaigns? We use data from the 2004, 2006,
and 2008 Pew Internet and American Life Post-Election Tracking Surveys
to examine the question of how individuals use the Internet in these elections
and to determine what makes an individual a political Internet user.
We take advantage of three surveys conducted by the Pew Internet and
Life Project in 2004, 2006, and 2008 to gain some insight into the differences
between individuals who are regularly using the Internet and those who
are not. These surveys allow us to look for differences in the characteristics
of Internet users compared to nonInternet users as well as to test for
differences in their political behavior.
3
Pew conducted these surveys via telephone interviews of adults living in
the continental United States using random digit dialing with both landline
phones and cell phones to produce a nationally representative sample. As
many as 10 attempts were made to contact each phone number, and phone
calls were staggered over days of the week and hours of the day. Each survey
was in the field for approximately two weeks after the November general
election and was able to obtain data on approximately 2,300 respondents:
November 2008 included 2,254 adults, 1,591 Internet users, and 1,186 online
political users, for example, whereas November 2004 included 2,200 adults,
of whom 1,324 were Internet users and 937 were online political users.
Each survey asks the respondents to identify the ways in which they use
the Internet; in this analysis, we will focus both on basic Internet usage as
62 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

well as Internet usage targeted for gaining information about or communicat-
ing about politics. We focus on whether there are statistically significant dif-
ferences between Internet users and non-Internet users in terms of their
socioeconomic and demographic variables as well as in their political prefer-
ences and opinions. We focus on patterns that we see emerging for each
year, and we address whether these patterns are different for 2008.
INTERNET ACCESS: CHANGES FROM 2004 THROUGH 2008
Before delving into the data on the use of Internet in American politics, it is
important to put the use of the Internet in general into context. The U.S.
Bureau of the Census has been asking American households about computer
usage since 1984 and about Internet usage since 1997 as part of the Current
Population Survey. As we see in Figure 1, since 2001, more than half of
American households have had connectivity to the Internet. However, between
2001 and 2007, Internet connectivity only increased by about 10 percentage
points; just more than 60 percent of households have an Internet connection.
In Table 1, we examine Internet connectivity in 2007 and see that there
remain gaps between those who are and are not connected. The better
educated are more likely to be online compared to those with high school
degrees or less; whites and Asians are also more likely to have home Internet
access compared to blacks and Hispanics. Table 1 also illustrates an interest-
FIGURE 1 Internet access in the United States, 19972007.
The American Internet Voter 63
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

ing point about studying the Internet and politics. When we compare the first
three columnswhich show household Internet access in 2007with the
fourth columnwhich are individual voting percentages in 2008we see
that there are often disconnects between households with Internet access
and people in those households voting. These disconnects are more appar-
ent when we compare the individual voting in 2008 by demographics with
the data on political Internet use in 2008 from the Pew Internet and American
Life Project data. Here, we see that these disconnects remain significant,
especially for older voters, who vote at high rates but are not political users
compared to younger individuals who claim to be political Internet users but
do not vote at nearly the same rate as they engage in online political activity.
This basic question about the relationship between Internet political use and
voting is a topic for further exploration, especially for populations with high
Internet penetration and low voting rates (e.g., younger individuals).
INTERNET USE IN THE POLITICAL PROCESS: CHANGES
FROM 2004 TO 2008
Our examination of changes in the use of the Internet in politics starts with a
descriptive review of data from the Pew Internet and American Life Project.
4
The changes in the use of the Internet in politics can be seen first in the
TABLE 1 Internet Access by Household, by Age, Race, and Education
Households
with
Internet
Households
with
broadband
Someone in
household
can access
the Internet
Voted
in 2008
Political
Internet
user 2008
% % % % %
Total 61.7 50.8 71.0 63.6 55.0
Younger than 25 57.7 51.9 74.5 48.5
2534 65.6 58.3 78.9 60.0 (>30) 72.0
3544 71.8 61.4 82.5 (3049) 65.0
4555 70.7 57.9 79.8 69.5 (5064) 51.0
55 and older 50.2 37.5 55.9 (65) 22.0
White non-Hispanic 66.9 54.9 75.1 66.1 58.0
Black 45.3 36.8 59.1 64.7 80.0
Asian 67.8 60.2 77.2 47.6

Hispanic (of any race) 43.4 35.2 54.8 49.9 52.0
Less than high
school graduate
24.0 17.1 32.4 39.4 18.0
High school graduate 49.5 36.8 59.1 54.9 42.0
Some college=associate
degree
68.9 56.0 79.5 68.0 67.0
Bachelors
degree=higher
84.0 73.9 90.6 78.9 81.0
Note. Source: U.S. Census Bureau, Current Population Survey, November 2007 and November 2008.

The data are missing.

64 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

differences in survey questions asked in 2004 compared to 2008. For
example, the 2008 survey contained questions about the use of Facebook
and Twitter, applications that were not widely used in 2004. The 2008 survey
also asks questions about blogging more directly than was done previously,
when it was included generally as a part of discussions of online forums and
Listservs. Below, we conduct paired sample t-tests on the means of a series of
variables where we compare Internet users to nonInternet users using the
2004 and 2008 data. It is particularly convenient to compare 2004 data to
2008 data, as both years represent presidential election contests, when
general interest about politics is likely to be particularly high.
2004 Survey Data
Full-time employment again was a statistically significant demographic charac-
teristic differentiating Internet and nonInternet users in 2004. Almost twice as
many Internet users as nonInternet users were fully employed. Overall, 51
percent of Internet users reported that they lived in suburban areas, but only
43 percent of nonInternet users said they lived in suburban communities.
Marriage rates between Internet and nonInternet users differed by about 10
percent, with more Internet users than nonInternet users being married.
Although 36 percent of Internet users and 35 percent of nonInternet
users considered themselves conservative, 45 percent of Internet users and
39 percent of nonInternet users voted for Bush. Similarly, 32 percent of
Internet users called themselves Republican but only 23 percent of non
Internet users categorized themselves as Republican. These statistics suggest
that, as a broad generalization, Internet users are slightly more Republican
than nonInternet users.
The differences between Internet and nonInternet users who follow
current political affairs most of the time show that Internet users are more
up-to-date with current politics than nonInternet users. NonInternet users
were less politically active than Internet users, too. Although statistically
insignificant, 10 percent fewer nonInternet users than Internet users were
registered to vote in November 2004. Internet usage also made no difference
in when respondents decided for whom to vote. These results suggest that
although the Internet may keep people better informed about current issues,
this does not influence how or whether Americans vote.
2008 Pew Internet and American Life Survey Data
By 2008, 55 percent of Internet users and 45 percent of nonInternet users
lived in suburbia. In addition, 57 percent of Internet users and only 38
percent of nonInternet users reported that they were married. The demo-
graphics of American Internet voters can be seen in Table 2. Full-time
employment is no longer statistically significant when comparing mean
The American Internet Voter 65
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

employment of the Internet and nonInternet user groups. Almost a majority
of both nonInternet and Internet users voted for Obama and Biden in the
2008 election; 48 percent of Internet users and 54 percent of nonInternet
users voted for Obama, but only 38 percent of Internet users and 27 percent
of nonInternet users voted for McCain (25 percent of survey respondents
refused to specify which candidate they voted for). Finally, 10 percent more
Internet users claimed to get their political information from Web sites that
share their point of view than from sites that challenge their point of view.
MULTIVARIATE SURVEY ANALYSIS
The previous section described only minimal differences between Internet
users and nonInternet users and noted that there were no dramatic differ-
ences in our analyses between the patterns of use in 2004 compared to
2008. In our multivariate analysis, we consider the factors that affect five
types of behaviors involving the Internet(1) general Internet access, (2)
accessing news online, (3) acquiring political information online, (4) using
e-government, and (5) sending and receiving political e-mails.
5
Our goal is
to examine the extent to which these behaviors are predicted by covariates
in our survey data, especially political variables. The political variables we
examine are (1) Democratic Party candidate choice, (2) party identification
(Democrat), and (3) campaign contact. The first two of these variables are
personal political attributes and the third attribute is a proxy for whether
the respondent had been exposed to other information about the campaign.
TABLE 2 Internet Access by Individual, by Age, Race, and Education
Households
with Internet
Households
with
broadband
Someone in
household
can access
the Internet
Voted in
2008
(Individuals)
% % % %
Total 61.7 50.8 71.0 63.6
Younger than 25 57.7 51.9 74.5 48.5
2534 65.6 58.3 78.9 60.0
3544 71.8 61.4 82.5
4555 70.7 57.9 79.8 69.5
55 and older 50.2 37.5 55.9
White non-Hispanic alone 66.9 54.9 75.1 66.1
Black alone 45.3 36.8 59.1 64.7
Asian alone 67.8 60.2 77.2 47.6
Hispanic (of any race) 43.4 35.2 54.8 49.9
Less than high school graduate 24.0 17.1 32.4 39.4
High school graduate 49.5 36.8 59.1 54.9
Some college, associate degree 68.9 56.0 79.5 68.0
Bachelors degree or higher 84.0 73.9 90.6 78.9
Note. Source: U.S. Census Bureau, Current Population Survey, November 2007 and November 2008.
66 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

To the extent each of these variables are predictive, we then gain some
insight into the degree to which Internet use may result in different out-
comes. In particular, we challenge the commonly held belief that particular
groups of partisans are more or less likely to have access to the Internet.
First, we generate a binary variable if the respondent reports any Inter-
net use. We then generate additional variables if the respondent reports get-
ting news online, getting political campaign information online, or getting
information from governmental Web sites online. Our final variable is an
indicator that describes whether the respondent sends or receives political
e-mails. We summarize each of our variables in Table 3. These variables
can be interpreted as a percentage (with a value of 1 being 100 percent),
so that for example in 2004, 31 percent of the respondents reported sending
or receiving political e-mails.
These variables are very closely comparable across years despite varia-
tions in survey wording. There are examples that make comparisons difficult:
in 2006, for example, the survey wording was quite different regarding
whether the respondent sent or received political e-mails, which makes it dif-
ficult to drawan actual comparison for e-mail across years. However, most sur-
vey questions are nearly identical. For most questions, the question wording is
sufficiently comparable that we are able to discern clear patterns in the respon-
dents Internet use patterns. Additionally, within each year, each question does
allow us to investigate the relationship between these variables and the
demographic and socioeconomic characteristics of the survey respondents.
Starting in 2004, we see that 60 percent of the respondents report having
Internet access. This increases to 66 percent in 2006 and to 70 percent in
2008. Only 42 percent of respondents report accessing their news online in
2004, where 50 percent report accessing their news online in 2008. The dif-
ference is particularly stark in terms of the 2004 to 2008 difference for respon-
dents who report getting their political information online: 39 percent in 2004
and 74 percent in 2008. We also see a slight increase in the percentage of
respondents who report accessing the Internet to gain information about
e-government: 34 percent in 2004 increased to 42 percent in 2008. There is
TABLE 3 Internet Usage Summary by Variable Mean: 2004, 2006, and 2008
Variable 2004 2006 2008
Internet usage 0.60 (0.49) 0.66 (0.47) 0.70 (0.46)
News online 0.42 (0.49) 0.40 (0.49) 0.50 (0.50)
Politics online 0.39 (0.49) 0.29 (0.46) 0.74 (0.44)
E-government 0.34 (0.47) 0.32 (0.47) 0.42 (0.49)
Sends=receives political e-mail 0.31 (0.46) 0.11 (0.31) 0.69 (0.46)
Total observations 2,197 2,562 2,254
Note. Source: PEW Internet and Life Project November 20042008 surveys. Specific question wording is
including in the appendix. Standard deviations are in parentheses below each mean. While these patterns
denote potentially interesting trends, note that no values are statistically distinguishable across years.
The American Internet Voter 67
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

a decrease in 2006 for the number of individuals who report sending and
receiving political e-mails in comparison to 2004 and 2008, perhaps because
of lower interest due to it being an off-year election.
In Table 4, we summarize each of the characteristics we observe for
each respondent. We are particularly interested in determining the extent
to which a respondents political affiliationsparty identification or vote
choicewill be different between Internet users and nonInternet users.
While most of these variables are again percentages, income and education
are categorical and age is reported in years.
Because of the breadth of the data, we are able to determine whether
the respondent recalls being contacted by a political campaign. We are also
able to ascertain whether the respondent believes the country is heading in
the right direction. We know a large number of demographic variables about
each respondent, including their age, marital status, employment status, party
identification, income, race, and gender. In each of the regressions, these
become key independent variables in our models. Our respondents were
slightly older in 2008 than in 2004 (49.4 years in 2004 compared to 55.2 years
in 2008) and have slightly more income. They are also less likely to believe
that the country is heading in a satisfactory direction.
Tables 5 through 9 present the coefficients from logistic regressions,
which incorporate each of the variables we observe for each respondent
as independent variables and try to predict the online behaviors. Each unit
of observation is an individual, while each column represents the coefficients
from that regression for this particular year. Our primary goal from these
exercises is to look for trends in prediction, that is, trends in participation
in online government or online politics in some way and, in particular, trends
in Internet use. We set the statistically significant coefficients in boldface so
that these patterns emerge. Because our dependent variable in each case is
binarywhether an individual engages or does not engage in a specific
online activitywe use logistic regression for each analysis.
TABLE 4 Independent Variable Summary by Year: 2004, 2006, and 2008
Variable 2004 2006 2008
Country heading in positive direction 0.46 0.31 0.16
Age 49.4 50.7 55.2
Education level 2.82 2.84 2.83
Married 0.57 0.55 0.56
Employed 0.47 0.48 0.41
Democratic ID 0.33 0.32 0.35
Income 4.79 4.93 5.96
Campaign contact 0.66 0.79 0.20
Non-white 0.19 0.18 0.17
Female 0.53 0.52 0.53
Total observations 2,197 2,562 2,254
Note. Source: PEW Internet and Life Project November 20042008 surveys.
68 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

Internet Access
Across each of the five models, the regression coefficients show a clear trend:
the role of the Internet in political life is not dominated by politics. Looking
first at Internet access in 2004 through 2008 (Table 5), we see that there are
TABLE 5 Logistic Regression Coefficients for Internet Access, 2004, 2006, and 2008
Variable 2004 Std Err 2006 Std Err 2008 Std Err
Democratic candidate choice 0.24 .166 0.04 .133 0.17 .151
Country heading in positive direction 0.11 .116 0.10 .116 0.57 .165
Democratic ID 0.03 .131 0.30 .128 0.13 .153
Campaign contact 0.64 .116 0.29 .132 0.06 .148
Age 0.05 .003 0.05 .003 0.06 .004
Education level 0.73 .062 0.76 .058 0.99 .067
Married 0.22 .116 0.12 .112 0.64 .123
Employed 0.29 .118 0.26 .113 0.83 .139
Female 0.7 .109 0.14 .104 0.24 .121
Income 0.21 .034 0.29 .032 0.05 .020
Non-white 0.54 .142 0.31 .136 0.67 .160
Missing age 3.05 .384 . . . . . . . . . . . .
Missing education level 1.24 .686 1.71 .613 3.04 .482
Missing income 0.441 .196 0.993 .187 0.98 .438
Constant 0.51 .272 0.55 .248 0.82 0.258
N 2,197 2,562 2,254
Pseudo R
2
0.27 0.28 0.34
Note. Values in bold indicate coefficients that are statistically significant at the 95% level. Dependent vari-
able is an indicator that describes whether the respondent has Internet access.
TABLE 6 Logistic Regression Coefficients for Accessing News Online, 2004, 2006, and 2008
Variable 2004 Std Err 2006 Std Err 2008 Std Err
Democratic candidate choice 0.21 .156 0.13 .118 0.22 .131
Country heading in positive direction 0.43 .112 0.11 .101 0.45 .142
Democratic ID 0.17 .127 0.20 .118 0.12 .135
Campaign contact 0.51 .112 0.62 .124 0.06 .127
Age 0.04 .003 0.03 .003 0.05 .003
Education level 0.70 .059 0.61 .052 0.88 .058
Married 0.18 .113 0.09 .101 0.19 .109
Employed 0.57 .112 0.33 .098 0.53 .109
Female 0.20 .104 0.00 .092 0.20 .105
Income 0.21 .032 0.19 .028 0.05 .019
Non-white 0.34 .138 0.07 .126 0.52 .141
Missing age 2.87 .426 . . . . . . . . . . . .
Missing education level 0.78 1.07 0.02 1.06 2.96 .465
Missing income 0.46 .204 0.76 .185 0.64 .289
Constant 1.46 .261 2.48 .240 0.36 .255
N 2,197 2,562 2,254
Pseudo R
2
0.21 0.17 0.25
Note. Values in bold indicate coefficients that are statistically significant at the 95% level. Dependent vari-
able is an indicator for whether the respondent accesses news online.
The American Internet Voter 69
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

differences across the three elections, especially between the off-year 2006
election and the two presidential elections. The constants across all three
elections are that older individuals use the Internet significantly less, that
better-educated individuals access the Internet more, and that individuals
TABLE 7 Logistic Regression Coefficients for Acquiring Political News and Information
Online, 2004, 2006, and 2008
Variable 2004 Std Err 2006 Std Err 2008 Std Err
Democratic candidate choice 0.21 .157 0.30 .124 0.32 .127
Country heading in positive direction 0.11 .113 0.12 .108 0.02 .135
Democratic ID 0.17 .129 0.22 .125 0.02 .130
Campaign contact 0.76 .115 0.35 .132 0.21 .126
Age 0.04 .004 0.03 .003 0.00 .003
Education level 0.75 .061 0.72 .058 0.06 .052
Married 0.25 .115 0.09 .109 0.38 .102
Employed 0.21 .112 0.06 .104 0.05 .110
Female 0.05 .106 0.28 .099 0.18 .100
Income 0.19 .032 0.17 .030 0.03 .018
Non-white 0.39 .141 0.23 .136 0.05 .139
Missing age 2.76 .418 . . . . . . . . . . . .
Missing education level 1.78 .803 . . . . . . 0.45 .486
Missing income 0.55 .209 0.56 .201 0.77 .284
Constant 1.95 .269 2.48 .258 0.58 .256
N 2,197 2,545 2,254
Pseudo R
2
0.22 0.16 0.02
Note. Values in bold indicate coefficients that are statistically significant at the 95% level. The dependent
variable is an indicator for whether the respondent acquires political news online.
TABLE 8 Logistic Regression Coefficients for Using E-Government, 2004, 2006, and 2008
Variable 2004 Std Err 2006 Std Err 2008 Std Err
Democratic candidate choice 0.21 .156 0.22 .120 0.25 .126
Country heading in positive direction 0.15 .115 0.17 .105 0.26 .136
Democratic ID 0.10 .129 0.11 .121 0.12 .130
Campaign contact 0.60 .115 0.76 .135 0.02 .126
Age 0.02 .003 0.02 .003 0.03 .003
Education level 0.67 .060 0.56 .055 0.72 .055
Married 0.16 .114 0.24 .104 0.20 .105
Employed 0.27 .113 0.35 .102 0.57 .104
Female 0.06 .105 0.08 .096 0.14 .100
Income 0.21 .032 0.16 .029 0.04 .019
Non-white 0.23 .142 0.14 .132 0.54 .138
Missing age 1.32 .409 . . . . . . . . . . . .
Missing education level 0.81 1.06 1.17 .795 2.04 .478
Missing income 0.41 .215 0.38 .200 0.38 .256
Constant 2.82 .281 3.32 .265 1.16 .252
N 2,197 2,562 2,254
Pseudo R
2
0.18 0.14 0.18
Note. Values in bold indicate coefficients that are statistically significant at the 95% level. The dependent
variable is an indicator for whether the respondent reports accessing e-government resources.
70 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

who are employed access the Internet more. When we look at the political
variables, we see that there is no effect of voting for the Democratic candi-
date or for identifying with the Democratic party. Only in 2008 is there any
effect for the direction of the country on Internet access. Campaign contact
is correlated with increased Internet access in 2004 and 2006 but not in
2008, and race (nonwhite) is negatively associated with Internet access in
2004 and 2008.
Accessing News and Acquiring Political Information Online
Table 6 shows that, when we consider who accessed online news in 2004
and 2006, older people access news less than young people. We also see that
education level matters in both the 2004 and 2006 elections, as does income.
Importantly, we do not see any explicitly political factors having an effect on
accessing the news online in 2004 or 2006. In 2008, there is a statistically sig-
nificant effect of decreased access of news online for individuals who report
that the country is heading in a positive direction.
When we look at acquiring political information online, we see the
same age effects as with accessing news online. We also see in Table 7 that
in 2004 and 2006, the individuals contacted by candidates also acquired more
political information than did others. Vote choice and party identification did
not affect the decision about acquiring political information online in 2004,
suggesting that the Internet is not purely the playground of one party or
the other. Yet in both 2006 and 2008, there was a statistically significant
TABLE 9 Logistic Regression Coefficients for Sending Political E-mails, 2004, 2006, and 2008
Variable 2004 Std Err 2006 Std Err 2008 Std Err
Democratic candidate choice 0.49 .158 0.30 .164 0.20 .121
Country heading in positive direction 0.30 .119 0.11 .151 0.27 .133
Democratic ID 0.01 .132 0.04 .168 0.08 .125
Campaign contact 0.75 .120 0.66 .217 0.37 .121
Age 0.03 .004 0.01 .004 0.02 .003
Education level 0.72 .063 0.72 .089 0.07 .051
Married 0.27 .117 0.05 .151 0.38 .101
Employed 0.22 .116 0.11 .146 0.39 .103
Female 0.12 .109 0.04 .136 0.04 .096
Income 0.18 .033 0.18 .043 0.01 .018
Non-white 0.65 .150 0.21 .198 0.19 .133
Missing age 1.82 .443 . . . . . . . . . . . .
Missing education level 1.21 1.07 . . . . . . 0.61 .426
Missing income 0.48 .221 0.27 .324 0.13 .241
Constant 2.78 .221 5.23 .428 0.22 .245
N 2,197 2,545 2,254
Pseudo R
2
0.19 0.12 0.04
Note. Values in bold indicate coefficients that are statistically significant at the 95% level. The dependent
variable is an indicator for whether the respondent reports sending political e-mails.
The American Internet Voter 71
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

and positive effect of voting for the Democratic presidential candidate
(Obama) and acquiring political information.
Using E-Government
In Table 8, we see a strong digital divide affect in 2004 and 2006but not a
political effectin those who use e-government. Age, education, income,
and employment all are strong predictors of using e-government. In 2008,
there was an indication that individuals who voted for Democratic candidates
were also more likely to be using e-government than were others.
Sending Political E-Mails
The seeking of information is different from the active use of the Internet for
expressing oneself or proactive political activities. We also see, across all
three elections, that individuals who were contacted by candidates or
political parties also engaged in more e-mailing than did others. This could
be the result of candidates contacting more high-propensity voters, who
would be more likely to engage in such activities. In 2004 we also see that
individuals who voted for Democrats were more likely to contact others.
To summarize our findings across each of these behaviors, we again
look for patterns of use. Individuals who use the Internet at all or for
accessing news online were less likely to do so if they were older (6=6 age
coefficients are negative and statistically significant) or nonwhite (5=6 coeffi-
cients are negative and statistically significant). Individuals are more likely to
use the Internet or to access news online if they were better educated (6=6
education coefficients are positive and statistically significant), had higher
incomes (6=6 coefficients are positive and statistically significant), were
employed (6=6 coefficients are positive and statistically significant), or were
contacted by the campaign (4=6 coefficients are positive and statistically sig-
nificant). The coefficients for the missing data variables are mixed (4=14
negative and statistically significant, 9=14 positive and statistically signifi-
cant). We observe one political coefficient that is significant: In 2008, indivi-
duals who report that the country is heading in a positive direction were less
likely to access news online.
We contrast these patterns with the variables used to describe the
political use of the Internet: accessing specifically political information, using
the Internet as an e-government resource, or sending and receiving political
e-mails. Here we again see less use associated with age (7=9 age coefficients
are negative and statistically significant, 1=9 age coefficients are positive and
statistically significant) or nonwhite race (4=9 negative and statistically signifi-
cant). We also see more use associated with education (7=9 coefficients are
positive and statistically significant), contact by the campaign (7=9 coeffi-
cients are positive and significant), and income (6=9 positive and statistically
72 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

significant). This pattern is less clear for employment (3=9 positive and
statistically significant, 1=9 negative and statistically significant) and for the
coefficients for missing data variables (3=21 negative and statistically signifi-
cant, 7=21 positive and statistically significant). We observe many more
political variables that are significant. Of the nine coefficients for the indicator
if the respondent has voted for the Democratic candidate, four are statistically
significant and positive. Twice the coefficient for the indicator whether the
country is heading in a positive direction is significant, but the coefficient is
once negative and once positive. Key to all of these regressions, however, is
that not a single time does the party identification indicator yield a statistically
significant coefficient.
CONCLUSION
We find that although there are observable variables that drive Internet use
and consumption of news and political information online, they are not
variables that are broadly characterized as political; instead, they follow
general patterns of those relating to the allocation of time. This suggests that
in fact any differences that we observe emerging from the American Internet
voter are due to exposure to the medium of the Internet itself and will allow
us in future research to investigate mechanisms that can drive differences in
policy preferences. We conclude with a single set of regression coefficients in
Table 10. Here, we look at two pairs of variables. First, we compare an
indicator variable for whether respondents reported turning out to vote with
their self-stated reason for going online: whether they went online to confirm
their own political views. Consistent with work by Diana Mutz (2006), those
who went online to confirm their views are in fact more likely to vote.
Additionally, again contrasting 2004 to 2008, we focus on individuals who
made political donations in those years and consider the relationship
between political donation and whether the respondent has Internet access.
Here, those with Internet access are more likely to make political donations.
TABLE 10 Implications of Online Participation, 2004, 2006, and 2008
Variable
(Std Err) Voted 2004 Voted 2006 Voted 2008
Political
donations
2004
Political
donations
2008
Went online
to confirm
views
1.17 (.234) 1.00 (.192) 1.30 (.233) . . . . . .
Uses Internet . . . . . . . . . 1.00 (.130) 3.54 (.714)
Constant 1.25 (.054) 0.69 (.043) 1.59 (.062) 2.19 (.112) 5.83 (.708)
N 2,197 2,562 2,254 2,197 2,254
Pseudo R
2
0.01 .01 0.02 0.03 0.08
Note. Values in bold indicate coefficients that are statistically significant at the 95% level.
The American Internet Voter 73
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

Worthy of note, however, is that the coefficient in 2008 was 3.5 times larger
than in 2004. The role of the Internet in individuals political lives is changing.
Yet, the basic American Internet voter has not yet become polarized by
party identification. Across each of our regressions, we see zero statistically
significant coefficients for Democratic identification. Other socioeconomic
or demographic variables that predict other types of political participation
as well, such as education and age, are good predictors of Internet activities.
NOTES
1. Of course, as Ambinder notes in the article, the well-functioning post office, newspapers, and radio
were also revolutions in politics as well.
2. For an example of such a study, see Arceneaux, Kevin, and David Nickerson. (2006). Even if you
have nothing nice to say, go ahead and say it: Two field experiments testing negative campaign tactics.
Unpublished manuscript. Prepared for presentation at the 2005 Meeting of the American Political Science
Association, September 14, Washington, DC.
3. The Pew Internet and American Life Project publishes an analytical report for each election. See
Smith 2009; Rainie and Smith 2008; Rainie, Horrigan, and Cornfield 2005.
4. Data from this project and related reports can be found at http://pewinternet.org/Reports/2009/6
The-Internets-Role-in-Campaign-2008.aspx (2008 report); http://pewinternet.org/Reports/2007/Election-
2006-Online.aspx (2006 report); and http://pewinternet.org/Reports/2005/The-Internet-and-Campaign-
2004.aspx (2004 report).
5. The specific questions used to code each of these categories are listed in the Appendix.
REFERENCES
Altman, M. and G. Klass. (2005). Current research in voting, elections, and
technology. Social Science Computer Review, 23(3), 269273.
Ambinder, M. (2008, June). HisSpace. The Atlantic. Retrieved ______ from http://
www.theatlantic.com/doc/200806/ambinder-obama.
Anderson, D. M. and M. Cornfield. (2003). The civic web: Online politics. Ranham,
MD: Rowman and Littlefield.
Best, Samuel J. and Brian S. Krueger. (2005). Analyzing the representativeness of
Internet political participation. Political Behavior, 27(2), 183216.
Bimber, B. (2003). Information and American democracy: Technology in the
evolution of political power. New York, NY: Cambridge University Press.
Bimber, B. and R. Davis. (2003). Campaigning online: The Internet in U.S. New York,
NY: Oxford University Press.
Chadwick, A. (2006). Internet politics: States, citizens, and new communication
technologies. New York, NY: Oxford University Press.
Cornfield, M. (2004). Politics moves online: Campaigning and the Internet.
New York, NY: The Century Foundation.
Foot, K. and S. Schneider. (2006). Web campaigning. Cambridge, MA: MIT Press.
Frantzich, S. (2003). Cyberage politics 101: Mobility, technology, and democracy.
New York, NY: Peter Lang Publishing.
Graff, G. (2007). The first campaign: Globalization, the Web, and the race for the
White House. New York: Farrar, Straus, and Giroux.
74 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

Hara, N. and Z. Estrada. (2005). Analyzing the mobilization of grassroot activities via
the Internet: A case study. Journal of Information Science, 31(1), 503514.
Kenski, Kate and Natalie Jomini Stroud. (2006). Connections between Internet use
and political efficacy, knowledge, and participation. Journal of Broadcasting
& Electronic Media, 50(2), 173192.
Kimball, David C. and Cassie A. Gross. (2006). The growing polarization of American
voters. In John C. Green and Daniel J. Coffey (Eds.), The state of the parties: The
changing role of contemporary American parties. 5th ed. Lanham, MD:
Rowman & Littlefield.
Klotz, R. (2004). The politics of Internet communication. Lanham, MD: Rowman and
Littlefield Publishers, Inc.
Klotz, R. (2007). Internet campaigning for grassroots and astroturf support. Social
Science Computer Review, 25(1), 311.
Krueger, Brian S. (2002). Assessing the potential of Internet political participation
in the United States: A resource approach. American Politics Research, 30(5),
476498.
Mossberger, K., C. Tolbert, and R. McNeal. (2008). Digital citizenship: The Internet,
society, and participation. Cambridge, MA: MIT Press.
Mossberger, Karen, Caroline Tolbert, and Mary Stansbury. (2003). Virtual inequality:
Beyond the digital divide. Washington, DC: Georgetown University Press.
Mutz, Diana C. (2006). Hearing the other side. Cambridge: Cambridge University
Press.
Norris, Pippa. (2001). Digital divide: Civic engagement, information poverty, and the
Internet worldwide. Cambridge: Cambridge University Press.
Rainie, Lee and Aaron Smith. (2008). The Internet and the 2008 election. Pew Inter-
net and American Life Project. Retrieved November 12, 2009, from http://
www.pewinternet.org/Reports/2008/The-Internet-and-the-2008-Election.aspx
Rainie, Lee, John Horrigan, and Michael Cornfield. (2005). The Internet and
campaign 2004. Pew Internet and American Life Project. Retrieved November
12, 2009, from http://www.pewinternet.org/Reports/2005/The-Internet-and-
Campaign-2004.aspx
Smith, Aaron, Kay Lehman Schlozman, Sidney Verba, and Henry Brady. (2009). The
Internet and civic engagement. Pew Internet and American Life Project.
Retrieved December 2, 2009, from http://www.pewinternet.org/Reports/2009/
15The-
Internet-and-Civic-Engagement.aspx
Smith, Aaron. (2009). The Internets role in campaign 2008. Pew Internet and
American Life Project. Retrieved November 12, 2009, from http://www.
pewinternet.org/Reports/2009/6The-Internets-Role-in-Campaign-2008.aspx
Tolbert, Caroline J. and Ramona S. McNeal. (2003). Unraveling the effects of the
Internet on political participation? Political Research Quarterly, 56(2), 175185.
Trent, J. S. and R. V. Friedenberg. (2007). Political campaign communication: Prin-
ciples and practices. Lanham, MD: Rowman & Littlefield.
Wagner, K. M. and J. Gainous. (2009). Electronic grassroots: Does online campaigning
work? The Journal of Legislative Studies, 15(4), 502520.
The American Internet Voter 75
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

APPENDIX: SURVEY QUESTIONS
Use the Internet:
1. Please tell me if you ever use your cell phone to do any of the following
things. Do you ever use it to access the Internet? (Q4b, 2008)
2. Do you use the Internet, at least occasionally? (Q6a, 2006; Q6a, 2008)
3. Do you ever go online to access the Internet or World Wide Web or to
send and receive e-mail? (Q6, 2004)
News Online:
1. Again, thinking about YESTERDAY, did you read a news story from a news-
paper in any of the following ways: Online on a computer? (Q12b, 2008)
2. Again, thinking about YESTERDAY, did you watch the news online on a
computer? (Q14b, 2008)
3. Please tell me if you ever use the Internet to do any of the following
things. Do you ever use the internet to. . . get news online? (WebA,
2006; Web1, 2008)
4. Please tell me if you do any of the following when you go online. Do you
ever get news online? Did you happen to do this yesterday, or not?
(Responses: Have ever done this, Did yesterday, Have not done this,
Dont know=Refused) (Web1, 2004)
Politics Online:
1. Did you ever go online to get news or information about the 2008
elections? (Q17, 2008)
2. Please tell me if you ever use the Internet to do any of the following
things. Do you ever use the Internet to. . . look online for news or
information about politics or the 2008 campaigns? (Web1, 2008)
3. How have you been getting most of your news about the November
elections . . . from television, from newspapers, from radio, from
magazines, or from the Internet? (Q15, 2008; Q19, 2006)
4. How have you been getting most of your news about the presidential
election campaign? From television, from newspapers, from radio, from
magazines, or from the Internet? (Q17, 2004)
5. Do you ever go online to get news or information about the 2004
elections? How often do you go online to get news about the elec-
tions . . . more than once a day, every day, three to five days per week,
one to two days per week, or less often? (Q19, 2004)
76 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

6. Please tell me if you do any of the following when you go online. Do you
ever look for news or information about politics and the campaign? Did
you happen to do this yesterday, or not? (Responses: Have ever done this,
Did yesterday, Have not done this, Dont know=Refused) (Web 1, 2004)
Access E-Government:
1. Please tell me if you ever use the Internet to do any of the following
things. Do you ever use the Internet to. . . visit a local, state, or federal
government Web site? (WebA, 2006; Web1, 2008)
2. Please tell me if you do any of the following when you go online. Do you
ever look for information from a local, state, or federal government Web
site? Did you happen to do this yesterday, or not? (Responses: Have ever
done this, Did yesterday, Have not done this, Dont know=Refused)
(Web 1, 2004)
Send=Receive Political E-mail:
1. Did you communicate with others about politics, the campaign, or the
2008 elections using the Internet, whether by e-mail, text messaging,
instant messaging or using a social networking site? (Q18, 2008)
2. Thinking about this years presidential election, people have been com-
municating with each other and with the political campaigns in many
ways, to talk about issues or where the campaign stands. What about
you? Over the past several months, how often did you. . . send or receive
E-MAIL to or from friends, family members, or others about the campaign?
(Q19, 2008)
3. Did you send or receive e-mails about the candidates or the campaigns,
either with personal acquaintances or political organizations, or did you
not happen to do this? (Q26, 2006)
4. Have you sent or received e-mails about the candidates or campaigns,
either with personal acquaintances or from groups or political organiza-
tions? (Q20, 2004)
5. Have you sent e-mails about the 2004 campaign to groups of family or
friends who are part of an e-mail list or online discussion group? (Q22, 2004)
6. During this years election campaigns, have you sent e-mails urging
people to get out and vote without reference to a particular candidate?
Sent e-mails urging people to vote for a particular candidate? (Q27,
2004)
7. When you went online to get information about the elections, did you
ever get or send e-mail with jokes about the campaigns and elections?
(Q37, 2004)
The American Internet Voter 77
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

Make Political Donations:
1. During this years election campaign, have you given money to a political
candidate? (Q27, 2004)
2. When you went online to get information about the elections, did you
ever contribute money online to a candidate running for public office?
(Q37, 2004)
3. There are many different campaign-related activities a person might do on
the Internet. Im going to read a list of things you may or may not have
done online in the months leading up to the November elections. Just tell
me if you happened to do each one, or not. Did you contribute money
online to a candidate running for public office? (Q26, 2006; Q25, 2008)
Was Contacted by Campaign (Other Than E-mail):
1. In the past two months, have you received mail urging you to vote for a
particular presidential candidate? Received telephone calls urging you to
vote for a particular presidential candidate? Been visited at home by
someone urging you to vote for a particular presidential candidate?
(Q23, 2004)
2. In the past two months, have you received mail urging you to vote for a
particular candidate? Been visited at home by someone urging you to vote
for a particular candidate? Received prerecorded telephone calls urging
you to vote for a particular candidate? Received a phone call from a live
person urging you to vote for a particular candidate? (Q2, 2006)
3. Thinking about this years presidential election, people have been com-
municating with each other and with the political campaigns in many
ways, to talk about issues or where the campaign stands. What about
you? Over the past several months, how often did you. . . Receive mail
from a candidate or political party? Receive text messages from a candi-
date or political party? (Q19, 2008)
Voted for the Democratic Candidate:
1. In the presidential election, did you vote for the Democratic ticket of Bar-
ack Obama and Joe Biden or the Republican ticket of John McCain and
Sarah Palin? (Vot03, 2008)
2. In the election on November 7, did you vote for the Republican candidate
or the Democratic candidate for Congress in your district? (Vot03, 2006)
3. In the election on November 2, did you vote for the Republican ticket of
George Bush and Dick Cheney, the Democratic ticket of John Kerry and
John Edwards, the ticket of Ralph Nader and Peter Camejo, or someone
else? (Vote 03b, 2004)
78 T. E. Hall and B. Sinclair
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4

Direction of Country:
1. Overall, are you satisfied or dissatisfied with the way things are going in
this country today? (Responses: Satisfied, Dissatisfied, (Not Read) Dont
know=Refused) (Q1, 2004; Q1, 2006; Q1, 2008)
Why Online:
1. When you go online looking for political or campaign information, would
you say most of the sites you go to share your point of view, dont have a
particular point of view, or challenge your own point of view? (Q41, 2004;
Q35, 2006; Q23, 2008)
Voted:
1. A lot of people have been telling us they didnt get a chance to vote in the
elections this year on November 4. How about you. . . did things come up
that kept you from voting, or did you happen to vote? (Vote, 2008)
2. A lot of people have been telling us they didnt get a chance to vote in the
congressional elections this year on November 7. How about you. . . did
things come up that kept you from voting, or did you happen to vote?
(Vot02, 2006)
3. A lot of people have been telling us they didnt get a chance to vote in the
elections this year on November 2. How about you. . . did things come up
that kept you from voting, or did you happen to vote? (Vot02, 2004)
The American Internet Voter 79
D
o
w
n
l
o
a
d
e
d

b
y

[
U
n
i
v
e
r
s
i
d
a
d

A
u
t
o
n
o
m
a

M
e
t
r
o
p
o
l
i
t
a
n
a
]

a
t

1
3
:
1
7

1
8

S
e
p
t
e
m
b
e
r

2
0
1
4